| Author | Zongwu Cai, Jiazi Chen, Linlin Niu |
| Content | It is well documented in the literature that individual saving decisions vary with the life cycle and at the macroeconomic level, a changing demographic age structure affects aggregated savings, which then drives a slow movement of interest rates. In this paper, we propose a semiparametric affine arbitrage-free yield curve model with a low-frequency trend structure driven by the entire age distribution through a life cycle impact function. The unified framework not only fully explores the demographic age structure to robustly explain yield trend, but also utilizes efficiently the interest rate term structure to infer the S-shaped age impact function of the whole life cycle. We estimate the model with quarterly U.S. data from 1950s to present. The results show clearly that the model fits U.S. Treasury yields remarkably well in sample and outperforms popular alternative models out of sample. After removing the demography-driven trend especially pertaining to the baby boomer’s life cycle, the remaining term structure component is stationary with counter-cyclical risk premia. |
| JEL-Codes | E43, G12, J11 |
| Keywords | Demographic distribution, Life cycle, Term structure models, Semiparametric model, Functional data analysis |
